| | 8 | I have finished my own testing of the integration of the ubercal zero points and flat-field values into the DVO system. The program 'setphot' loads the zero points and imports the flat-field corrections into an internal table, and sets the images to use the zero points and applies the zero points and flat-field values to the detections. The program 'relphot' determines relative photometry for the non-ubercal images using all overlaps, keeping the ubercal image zero points fixed. The test suite generates a fake set of images, zero points and flat-field corrections, ingests into dvo, then checks that setphot and relphot get the expected answers. This all worked quite well (after fixing bugs). I then applied the ubercal corrections to a subset of the 3pi dvo database, and extracted a sample of detections after the corrections were applied. I sent those off to Eddie for comparison to the ubercal'ed detections. The results look mostly quite good (to floating point precision), but there are a small number of ~0.5mmag discrepancies that we need to track down. |
